carbcor

Python package to calculate Placco et al. (2014) carbon abundance evolutionary corrections with online tool at vplacco.pythonanywhere.com.

Corrects observed [C/Fe] abundances for the depletion of carbon due to CN processing on the upper red giant branch. Inputs are observed [C/Fe], logg, and [Fe/H].

Installation

pip install carbcor

Quick start

Python API

from carbcor import get_carbon_correction, apply_carbon_corrections
import pandas as pd

# Single star
res = get_carbon_correction(logg=0.88, feh=-4.82, cfe=0.50) # these are the parameters for star GDR3_526285 [Limberg, Placco et al. (2025)](https://ui.adsabs.harvard.edu/abs/2025ApJ...989L..18L/abstract)
print(res["cfe_corrected"])   # corrected [C/Fe]
print(res["correction"])      # correction delta

# Batch from a DataFrame
# You can test for the whole JINAbase [Abohalima & Frebel 2018](https://ui.adsabs.harvard.edu/abs/2018ApJS..238...36A/abstract) following the tests folder.
# Since it parses 1 star at a time, it takes ~10min for each 2.5k stars.
df = pd.DataFrame({
    "logg": [0.88, -4.82, 0.50],  # GDR3_526285  [Limberg, Placco et al. (2025)](https://ui.adsabs.harvard.edu/abs/2025ApJ...989L..18L/abstract)
    "feh":  [1.89, -4.11, 2.20],  # HE 0144-4657 [Placco, Limberg et al. (2026)](https://ui.adsabs.harvard.edu/abs/2026ApJ..1000L..34P/abstract)
    "cfe":  [2.02, -4.06, 1.53],  # GDR3 47959   [Limberg, Placco et al. (2026)] --> submitted
})
df_cor = apply_carbon_corrections(df)
print(df_cor[["cfe", "cfe_corrected", "correction"]])

Command line

# Query a single star
carbcor query 0.88 -4.82 0.50 # GDR3_526285

# Process a CSV file
carbcor batch stars.csv -o corrected.csv

# Inspect the web form structure
carbcor discover

Parameter space

The underlying correction grid covers:

Values outside these ranges are clipped to the nearest grid point.

Custom column names

apply_carbon_corrections(df, logg_col="logg", feh_col="Fe/H", cfe_col="CFe") # example for the JINAbase_2021.csv table

How it works

1. The package carbcor fetches the website HTML and auto-discovers the form structure (action URL, HTTP method, field names).
2. For each star, it submits the parameters and parses the correction delta from the HTML response.
3. Corrected [C/Fe] = input [C/Fe] + correction delta.
4. Missing / NaN inputs produce NaN outputs (standard float('nan'), which is to be compatible with pandas, i.e., you can just plot it afterwards).

Development

git clone https://github.com/guilimberg/carbcor.git
cd carbcor
pip install -e ".[dev]"
pytest

Reference

If you use this tool, you SHOULD cite:

Placco, V.M., Frebel, A., Beers, T.C., & Stancliffe, R.J. 2014, ApJ, 797, 21 (ADS | arXiv:1410.2223)

License

MIT